We have demonstrated that CMV/SIV vectors can 1) re-infect CMV+ rhesus macaques (RM), 2) during re-infection, elicit potent and persistent SIV-specific CD4+ and CD8+ T cell responses with a strong effector memory (TEM) bias, and 3) completely protect ~50% of vaccinated RM from progressive infection after limiting dose rectal challenge with the highly pathogenic SIVmac239 virus. The protection manifested in these RM is distinct from previous vaccines in its abruptness and extent, with protected RM exhibiting a viral burst in plasma of varying size upon initial infection, followed by immediate control to undetectable levels. Although occasional viral blips in plasma are observed in protected RM, these decline with time, and after 1 year, protection is unaffected by CD8+ or CD4+ cell depletion, and extensive tissue analysis with ultrasensitive nested PCR has shown only rare detection of ~ single copy SIV nucleic acid and no viable SIV. Protection correlates with the total SIV-specific CD8+ TEM generated during the vaccine phase, and occurs without an anamnestic response. These data indicate a novel pattern of protection consistent with very early control, likely taking place at the site of viral entry and/or early sites of viral replication and amplification, and involving tissue-resident CD8+ TEM. Thus, CMV vectors and the TEM vaccine concept offer a powerful new approach to HIV/AIDS vaccine development. However, fully replicative CMV vectors are unlikely to be advanced to human use due to the pathogenic potential of CMV. A central goal is therefore to develop CMV vectors that maintain immunogenicity and efficacy, but are safe to use in humans. Fortunately, preliminary data indicates that replication/spread-deficient CMVs can be robustly immunogenic. Here, we will explore the extent to which replication/spread-deficient CMV vectors can be attenuated with respect to pathogenicity and transmissibility, while retaining immunogenicity and efficacy. Three specific aims are proposed: 1) to design, construct and in vitro characterize replication/spread-deficient CMV/SIV vectors, 2) to determine the pathogenicity of replication/spread-deficient CMV/SIV vectors in fetal RM, and 3) to determine the immunogenicity and shedding of replication/spread-deficient CMV/SIV vectors in juvenile RM, and to take the optimal replication/spread deficient vector design to an efficacy trial to determine its ability to protect against limiting dose rectal SIVmac239 challenge.